Dimitar Kitanovski

Mr. Dimitar Kitanovski

orcid

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
10 October 2025 in Data in Brief
Verified
Journal Article
Open access
JDJovana Dobreva
Jovana Dobreva
IKIvana Karasmanakis
Ivana Karasmanakis
FIFilip Ivanisevic
Filip Ivanisevic
THTadej Horvat
Tadej Horvat
DKDimitar Kitanovski
Dimitar Kitanovski
MSMonika Simjanoska
Monika Simjanoska
The paper introduces RAGCare-QA, an extensive dataset of 420 theoretical medical knowledge questions for assessing Retrieval-Augmented Generation (RAG) pipelines in medical education and evaluation settings. The dataset includes one-choice-only questions from six medical specialties (Cardiology, Endocrinology, Gastroenterology, Family Medicine, Oncology, and Neurology) with three levels of complexity (Basic, Intermediate, and Advanced). Each question is accompanied by the best fit of RAG implementation complexity level, such as Basic RAG (315 questions, 75.0 %), Multi-vector RAG (82 questions, 19.5 %), and Graph-enhanced RAG (23 questions, 5.5 %). The questions emphasize theoretical medical knowledge on fundamental concepts, pathophysiology, diagnostic criteria, and treatment principles important in medical education. The dataset is a useful tool for the assessment of RAG- based medical education systems, allowing researchers to fine-tune retrieval methods for various categories of theoretical medical knowledge questions.

22 January 2026 in Mathematics
Verified
Journal Article
Open access
PDF file
PSPetar Sekuloski
Petar Sekuloski
DKDimitar Kitanovski
Dimitar Kitanovski
IGIgor Goshev
Igor Goshev
KMKostadin Mishev
Kostadin Mishev
MSMonika Simjanoska
Monika Simjanoska
VRVesna Dimitrievska Ristovska
Vesna Dimitrievska Ristovska
Large language models (LLMs) have become central to modern artificial intelligence, yet their internal decision-making processes remain difficult to interpret. As interest grows in making these models more transparent and reliable, topological data analysis (TDA) has emerged as a promising mathematical approach for exploring their structure. This scoping review maps the current landscape of research where TDA tools—such as persistent homology and Mapper—are used to examine LLM components like attention patterns, latent representations, and training dynamics. By analyzing topological features across layers and tasks, these methods provide new ways to understand how language models generalize, respond to unfamiliar inputs, and shift under fine-tuning. The review also considers how TDA-based techniques contribute to broader goals in interpretability and robustness, especially in detecting hallucinations, out-of-distribution behavior, and representational collapse. Overall, the findings suggest that TDA offers a rigorous and versatile framework for studying LLMs, helping researchers uncover deeper patterns in how these models learn and reason.

1 September 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
LJLisa Jusufi
Lisa Jusufi
RARoberta Apostolska
Roberta Apostolska
DKDimitar Kitanovski
Dimitar Kitanovski
Eftim ZdravevskiEftim Zdravevski
Eftim Zdravevski
This paper presents the Skopje School Exposure Dataset, an exposure dataset comprising 198 school buildings, including 165 primary and 33 secondary schools, located in the Skopje region of North Macedonia. Developed as part of a broader nationwide exposure model for school buildings, the dataset provides general information for each asset, including its name, address, geographical location, and administrative classification. It also includes structural characteristics classified according to the GEM Building Taxonomy v3.2, such as construction material, lateral load resisting system, ductility level, and number of storeys. Additional information covers occupancy during school hours, including pupils and staff, as well as total floor area and estimated structural replacement cost. The dataset was compiled using a bottom-up approach that integrates multiple sources, including previous research projects, governmental and cadastral records, and publicly available online sources. The Skopje region is home to one third of the country's population, and therefore a large share of its school buildings and pupils, making it a priority for seismic risk assessment. The school buildings in the Skopje region cover approximately 534 thousand m² of floor area (~34% of the national total), with an estimated replacement cost of €433 million (~34% of the national total), and a combined occupancy of around 90.7 thousand occupants (~32% of the national total) across 17 municipalities. As the first openly available and harmonised exposure inventory of school buildings in the Skopje region, the dataset provides a transparent basis for seismic risk assessment, emergency preparedness planning, and the prioritisation of retrofit interventions for educational infrastructure.

25 September 2026 in Journal of Big Data
Verified
Journal Article
Open access
PDF file
DKDimitar Kitanovski
Dimitar Kitanovski
Paulo Jorge CoelhoPaulo Jorge Coelho
Paulo Jorge Coelho
Ivan Miguel PiresIvan Miguel Pires
Ivan Miguel Pires
Petre LameskiPetre Lameski
Petre Lameski
ICIvan Chorbev
Ivan Chorbev
Eftim ZdravevskiEftim Zdravevski
Eftim Zdravevski
The rapid growth in data volume, complexity, and heterogeneity has intensified the need for automated and scalable data design patterns and architectures s

9 October 2026 in Data in Brief
Verified
Journal Article
Open access
LJLisa Jusufi
Lisa Jusufi
RARoberta Apostolska
Roberta Apostolska
DKDimitar Kitanovski
Dimitar Kitanovski
EZEftim Zdravevski
Eftim Zdravevski
This paper presents the Skopje School Exposure Dataset, an exposure dataset comprising 198 school buildings, including 165 primary and 33 secondary schools, located in the Skopje region of North Macedonia. Developed as part of a broader nationwide exposure model for school buildings, the dataset provides general information for each asset, including its name, address, geographical location, and administrative classification. It also includes structural characteristics classified according to the GEM Building Taxonomy v3.2, such as construction material, lateral load resisting system, ductility level, and number of storeys. Additional information covers occupancy during school hours, including pupils and staff, as well as total floor area and estimated structural replacement cost. The dataset was compiled using a bottom-up approach that integrates multiple sources, including previous research projects, governmental and cadastral records, and publicly available online sources. The Skopje region is home to one-third-of the country's population, and therefore a large share of its school buildings and pupils, making it a priority for seismic risk assessment. The school buildings in the Skopje region cover approximately 534 thousand m² of floor area (∼34% of the national total), with an estimated replacement cost of €433 million (∼34% of the national total), and a combined occupancy of around 90.7 thousand occupants (∼32% of the national total) across 17 municipalities. As the first openly available and harmonised exposure inventory of school buildings in the Skopje region, the dataset provides a transparent basis for seismic risk assessment, emergency preparedness planning, and the prioritisation of retrofit interventions for educational infrastructure.